使用 tf.data 加载 NumPy 数据

在 Tensorflow.org 上查看 在 Google Colab 运行 在 Github 上查看源代码 下载此 notebook

本教程提供了将数据从 NumPy 数组加载到 tf.data.Dataset 的示例 本示例从一个 .npz 文件中加载 MNIST 数据集。但是,本实例中 NumPy 数据的来源并不重要。

安装


import numpy as np
import tensorflow as tf
import tensorflow_datasets as tfds

.npz 文件中加载

DATA_URL = 'https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz'

path = tf.keras.utils.get_file('mnist.npz', DATA_URL)
with np.load(path) as data:
  train_examples = data['x_train']
  train_labels = data['y_train']
  test_examples = data['x_test']
  test_labels = data['y_test']

使用 tf.data.Dataset 加载 NumPy 数组

假设您有一个示例数组和相应的标签数组,请将两个数组作为元组传递给 tf.data.Dataset.from_tensor_slices 以创建 tf.data.Dataset

train_dataset = tf.data.Dataset.from_tensor_slices((train_examples, train_labels))
test_dataset = tf.data.Dataset.from_tensor_slices((test_examples, test_labels))

使用该数据集

打乱和批次化数据集

BATCH_SIZE = 64
SHUFFLE_BUFFER_SIZE = 100

train_dataset = train_dataset.shuffle(SHUFFLE_BUFFER_SIZE).batch(BATCH_SIZE)
test_dataset = test_dataset.batch(BATCH_SIZE)

建立和训练模型

model = tf.keras.Sequential([
    tf.keras.layers.Flatten(input_shape=(28, 28)),
    tf.keras.layers.Dense(128, activation='relu'),
    tf.keras.layers.Dense(10, activation='softmax')
])

model.compile(optimizer=tf.keras.optimizers.RMSprop(),
                loss=tf.keras.losses.SparseCategoricalCrossentropy(),
                metrics=[tf.keras.metrics.SparseCategoricalAccuracy()])
model.fit(train_dataset, epochs=10)
Epoch 1/10
938/938 [==============================] - 2s 2ms/step - loss: 3.1713 - sparse_categorical_accuracy: 0.8769
Epoch 2/10
938/938 [==============================] - 2s 2ms/step - loss: 0.5085 - sparse_categorical_accuracy: 0.9271
Epoch 3/10
938/938 [==============================] - 2s 2ms/step - loss: 0.3764 - sparse_categorical_accuracy: 0.9466
Epoch 4/10
938/938 [==============================] - 2s 2ms/step - loss: 0.3165 - sparse_categorical_accuracy: 0.9550
Epoch 5/10
938/938 [==============================] - 2s 2ms/step - loss: 0.2812 - sparse_categorical_accuracy: 0.9599
Epoch 6/10
938/938 [==============================] - 2s 2ms/step - loss: 0.2587 - sparse_categorical_accuracy: 0.9645
Epoch 7/10
938/938 [==============================] - 2s 2ms/step - loss: 0.2530 - sparse_categorical_accuracy: 0.9674
Epoch 8/10
938/938 [==============================] - 2s 2ms/step - loss: 0.2192 - sparse_categorical_accuracy: 0.9707
Epoch 9/10
938/938 [==============================] - 2s 2ms/step - loss: 0.2116 - sparse_categorical_accuracy: 0.9721
Epoch 10/10
938/938 [==============================] - 2s 2ms/step - loss: 0.2014 - sparse_categorical_accuracy: 0.9747

<tensorflow.python.keras.callbacks.History at 0x7fe4f37d1470>
model.evaluate(test_dataset)
157/157 [==============================] - 0s 2ms/step - loss: 0.5586 - sparse_categorical_accuracy: 0.9568

[0.5586389303207397, 0.9567999839782715]